Jump to content

IntrepidWoT

Verified Tanker [NA]
  • Content Count

    226
  • Joined

  • Last visited

About IntrepidWoT

  • Rank
    Has Cheese with Whine

Profile Information

  • Server
    NA
  1. That's just crazy talk. You can't ask people to counter by like, driving around and coordinating an attack. You need an asymmetrical vehicle class that can fire over terrain and to nullify all the experience someone has with their tank and map, with the skill to execute. The next thing you know you'll be asking WoT Map designers to pretend a map can be designed not using a quadrant/chokepoint model and be fun.
  2. Ah, well, hopefully the results and work put into these two recent threads (one here, one on the official forums) compares favorably to those. I'm fairly certain a handful of the posters over on the official forums have an extra chromosome/are the result of multi-generational incest.
  3. I didn't realize, if you have a link I'd love to credit the results in the thread.
  4. http://forum.worldoftanks.com/index.php?/topic/355727-higher-skilled-players-like-artillery-less-confirmed-but-why/#topmost Using only official forum poll results. Of course now I'm lying/bastard/manipulating data/using obscure file formats/not using someone's pet non-linear regression analysis method/poll questions were wrong/weights are wrong/sample size too small etc. etc. etc. Some of the glue eating horde replying in that thread are so fucking dense it's hilarious.
  5. Well, on the off chance official forum moderation doesn't decide to close and remove the thread (I took great care to make it as simple and inoffensive as possible), I'll have numbers from the official forums as well. Depending on the quantity of votes, I may need help from someone I can trust gathering their recent WN8 data. Unlike this forum, I can't just go through each person's profile and easily see their recent WR and WN8. Any volunteers to assist would be greatly appreciated. Worst case I'll do it on my own.
  6. I have no confidence at all that the official forum moderators would allow that poll. What do they win by letting it stay up there? The content would become a flame fest almost overnight. The results would either lend further proof that artillery are the heroes of bads everywhere and they'd have to ignore it openly or it would show that's not the case in which case, they still have no reason to do anything differently. They already have excuses to not do anything differently... so allowing that poll would not benefit them in any way. I can try it, but I suspect I'll end up with an RO over there at the minimum. As for the quality of the conclusion, I feel good about it even with just this limited data. If you think of the effect bads listening to goods would have on the conclusion, it should be that it makes it impossible to predict opinions because there is little difference between a good/bad opinion on artillery. Instead, we find even with the effect of less skilled players listening to the better players regarding artillery, their difference in opinion is still enough to be measurable. What you'd likely see in an official forum opinion is that the bottom of that predicted line would go much lower, below a 6 down to maybe a 3 or a 2, but the peak would remain unchanged. I'm going to debate whether or not I should attempt it there. I haven't shared this with the OF and have no plans to at the moment. That wasn't the goal.
  7. For recent winrate, the correlation is even stronger. Note that there appears to be fewer points of data, but this isn't the case. I just didn't have a way to look beyond the forum's left hand box and that only provides whole numbers to work with. Because of that, you only see a point of data per whole number rather than a continuous stream of winrate percentages. The p-value here was .000000656. Again the numbers are such that they wouldn't predict anyone is content with artillery, but the more skilled you are (based on recent winrate) the more likely it is that you strongly dislike artillery and want them removed.
  8. Thinking through ways to prove this. While it would be relatively trivial to point out players with large numbers of artillery games and TD games, there is always the chance the TDs came first or concurrently with the artillery games. I could focus on TDs that came out only after 8.6, but that wouldn't make the broader case and would only, at best, show artillery players like those TDs. It may be that the best way to confirm that suspicion is just to compare artillery players vs. players actually playing WoT and point out how much more likely artillery players are to play large numbers of games in TDs. This would show at a minimum that TDs are the refuge of disgruntled clickers.
  9. Will put the charts up tomorrow, but it turns out recent Winrate is even more strongly correlated than recent WN8. To address critics, also ran it for Battle count and Avg. Tier to help remove the "well duh, because of the % of votes no matter what you check it's going to correlate" objection. Battles fought and Avg. Tier are not strongly correlated with artillery opinions. Still have to run it for overalls.
  10. I looked at the account of all voters, taking down their recent WN8, recent WR, overall WN8, overall WR, average tier, battle count and location. I've only taken the time to run one regression analysis on the numbers so far, WN8 against vote. To make this work, I assigned a numeric value to each of the voting options for artillery. To help show the spread of opinion, I made it a scale from 1-10, assigning values as follows: 10 = Fuck artillery and anyone who plays them 9 = Remove artillery 8 = Overhaul, but remove artillery if that's not enough 6 = Overhaul, but keep artillery 3 = Modify artillery, but no overhaul needed 2 = Artillery is fine 1 = I'm a clicker I grouped the top three and bottom three together. I went back and forth on the value for overhaul but keep, and opted for it being a 6 instead of a 5 based on the severity of changes that would be required, making it almost unrecognizable to its current form. The regression analysis showed that there is a statistically significant relationship between your WN8 and how you feel about artillery. Incredibly so (.000224 p-value). For those who like pictures, here's the chart. Each of those dots is the recent WN8 of a voter. The orange line shows the predicted vote based on WN8. Notice that it's sloping upward. Also notice that it wouldn't predict anyone was ok with artillery no matter how bad they are at the game. It starts at 6, "overhaul but keep". If you take away EU votes, it's even more lopsided. EU players are more tolerant of artillery at higher skill levels than NA players. The two little dots for "I love artillery" that are unicums are both EU as are the majority of super unicums voting 6 or less. I can't speak to SEA because I didn't have enough SEA votes to make any conclusion. I'll run recent winrate and others soon, but that was enough of a confirmation to let go of my worrying that it's just confirmation bias. So, not only did 73.22% recommend either removing or, if unable after a major overhaul to fix them removing, artillery, but the better you are at WoT, the more likely you are to strongly dislike artillery.
  11. Busier than expected in the period I'd normally reserve for something like this. May be a few days later than expected, 170+ people to take a look at. Will get through it by the end of the week though.
  12. PM me about server resources. I can't promise anything, but I may be able to help.
×
×
  • Create New...